§ Claim under review · Capability
"Sam Altman, criador do ChatGPT, disse que 'o mundo deve aceitar algumas coisas negativas em troca dos benefícios da IA'" (English: "Sam Altman, creator of ChatGPT, said that 'the world should accept some negative things in exchange for the benefits of AI'")
Verdict
Mostly accurate
Confidence
HighSummary
Sam Altman really did say this. In an interview published on 4 and 5 October 2026 with POLITICO's new Decoded newsletter and podcast, the OpenAI chief executive said that one of the biggest differences between his company and stricter AI safety advocates is that "we believe that the world should accept some bad things happening for the benefits of this technology and people having the agency". The fact-checking site Snopes verified the quote against the audio recording and confirmed it was correctly attributed. The post's caption is also fair: Altman drew a clear line at what he called truly catastrophic risks, naming a serious loss of control over AI as something he would not accept, and he argued that severely restricting public access to AI would carry its own harms. The only small imprecision is that the Portuguese sentence shown in quotation marks is a shortened translation rather than his exact words, leaving out the part about people having agency. Describing Altman as the creator of ChatGPT is loose shorthand, since he leads the company that built it rather than having built it himself. Nothing important about the statement is in dispute.
The readings
key figures from the evidencetimestamp in podcast where quote occurs, per Snopes
length of the Decoded podcast episode containing the quote
Why this verdict
Evidence
The statement is real and the attribution is correct. POLITICO published an article on 10/04/2026 headlined "Sam Altman to Decoded: 'The world should accept some bad things happening' for the benefits of AI", reporting that Altman said in an exclusive interview with Decoded by POLITICO that the technology's benefits justify accepting some risks and that AI should remain broadly accessible to the public . Snopes, which examined the claim after it spread online, concluded that Altman did produce the statement during the interview with Politico's The Decoded Podcast and rated the quote as correctly attributed to him . Snopes places the exchange at the 8:17 mark of the Oct. 5 episode, with Altman responding to a question about how he would enforce AI rules compared with Anthropic CEO Dario Amodei .
The verbatim English wording, as reproduced in the Snopes transcript excerpt, is: "I think one of the biggest differences between us and some of the stricter, let's say, AI safety people is we believe that the world should accept some bad things happening for the benefits of this technology and people having the agency" .
The caveat the Instagram caption reports is also in the record. Altman later added that he does not accept "the really catastrophic risks," including the potential for "a serious loss of control to AI" . He also said he would not take a trade guaranteeing no major hacks, no misuse and zero scams, because he believes people will do orders of magnitude more good than bad with the technology . Fortune reported that Altman said "catastrophic risks" are unacceptable, and that restricting access to AI too severely in the name of safety could itself be harmful .
The caption's framing about recent OpenAI and Anthropic convergence on safety is also supported. Snopes notes that in its article about the episode, Politico reported Altman agreed with Amodei on the need to slow down AI development and that Altman's lobbyists have been backing new AI safety requirements from the government .
Findings
✓ What's accurate 6
- Sam Altman did make the statement. The quoted idea is a faithful rendering of what he said in an interview with POLITICO's Decoded, published 4 and 5 October 2026.
- The English words he used were that the world should accept some bad things happening for the benefits of the technology and for people having agency. He framed this as a difference between OpenAI and stricter AI safety advocates, in answer to a question about Anthropic and Dario Amodei.
- The caption's caveat is accurate. Altman did exclude what he called the really catastrophic risks, naming a serious loss of control to AI as the line he would not cross.
- The caption's statement that Altman argued AI should remain broadly accessible to the public matches POLITICO's own summary of the interview.
- The caption's framing of a recent OpenAI and Anthropic convergence on safety and regulation matches reporting that Altman agreed with Amodei on slowing AI development and that OpenAI has backed some new government safety requirements.
- Describing Altman as CEO and co-founder of OpenAI is correct.
≈ What's misleading 3
- Quote manipulation, at a minor level: the Portuguese sentence inside quotation marks is a translation and a slight compression, not a word for word rendering. Altman said "we believe that the world should accept some bad things happening for the benefits of this technology and people having the agency". The post's version drops "we believe", drops the clause about people having agency, and renders "for the benefits of" as "in exchange for". None of these change the substance of what he argued, but readers should know the quotation marks enclose a translated paraphrase rather than his exact words.
- Omitted qualifier, at the headline level only: the headline sentence standing alone carries none of the limits Altman placed on the statement. The post's own caption restores them, quoting the "truly catastrophic" and "serious loss of control" exclusions, so the post as a whole is not misleading. The risk is that the headline sentence travels without the caption attached.
- Describing Altman as "criador do ChatGPT", the creator of ChatGPT, is journalistic shorthand rather than a precise description. He leads the company that built ChatGPT. He is not its individual creator.
? What's uncertain 1
- Nothing material to the verdict. The attribution, the wording, the context and the caveat all line up with the primary publisher and an independent transcript excerpt.
Sources
8 of 8 linked to recordsPOLITICO, "Sam Altman to Decoded: 'The world should accept some bad things happening' for the benefits of AI", dated 10/04/2026 4:55 PM EDT, the outlet that conducted the interview
The Decoded Podcast (POLITICO), episode "Sam Altman on rogue AI, Anthropic, and getting money out of politics", published 2026-10-05, 28 minutes, hosted by Brendan Bordelon
Snopes fact check, "Sam Altman said world needs to accept 'some bad things happening' for benefits of AI", published Oct. 8, 2026, containing a transcript excerpt beginning at the 8:17 mark
Fortune, 2026-10-05
Fox Business, 2026-10-05
American Bazaar, 2026-10-05, carrying the longest contiguous block of surrounding quotation found
Xataka Brasil and Gizmodo Brasil, Portuguese-language coverage of the same interview
POLITICO official X account, 2026-10-05, promoting the interview